RODRIGUEZ v. WEST PUBLISHING CORP.

OPINION

RYMER, Circuit Judge:

West Publishing Corp. and Kaplan, Inc. entered a settlement agreement in an antitrust class action brought by those who purchased a BAR/BRI course between August 1, 1997 and July 31, 2006. (BAR/BRI is a subsidiary of West that provides preparation courses for state bar exams.) The district court approved the settlement, and several class members who object (Objectors) appeal. Their principal objection relates to incentive agreements that were entered into at the onset of litigation between class counsel and five named plaintiffs who became class representatives. They also contend that the district court improperly failed to compare the amount of the settlement to the likely recovery of treble, as well as single, damages.

We agree that the ex ante incentive agreements created conflicts among the five contracting class representatives, their counsel, and the rest of the class. We disapprove of them. Nevertheless, there were two other class representatives who had no incentive agreements and whose separate counsel were not conflicted. They provided adequate representation and the court was not required to reject the settlement on this account.

We conclude that the district court did not clearly abuse its discretion in finding that the $49 million settlement was fair, adequate, and reasonable even though it evaluated the monetary portion of the settlement based only on an estimate of single damages. Courts are not precluded from comparing the monetary component of a settlement to the estimated treble damages if, in their informed judgment, the strength of the particular case warrants it; but they are not obliged to do so in every antitrust class action. In this case, the settlement is substantial and meets the standard for approval by any measure.

Finally, we believe that the incentive agreements may have an effect on attorney's fees that the district court did not acknowledge. It gave no weight to the Objectors' role in securing denial of incentive awards, nor did the court take into account ethics concerns arising out of the incentive agreements when it awarded attorney's fees to class counsel. Both issues need to be revisited.

The Objectors' remaining arguments lack force. Accordingly, we affirm approval of the settlement. We reverse the orders denying any fee award to Objectors and granting the fee award to class counsel, and remand.

I

Ryan Rodriguez and Reena B. Frailich brought this action on behalf of themselves and "[a]ll persons who purchased a bar review course from BAR/BRI in the United States from August 1997 to the present" against West and Kaplan. They filed a first amended complaint in May 2005 joined by Loredana Nesci, Jennifer Brazeal, and Lisa Gintz. Kari Brewer and Lorraine Rimson were named plaintiffs in a related action (Brewer v. West Publishing Corp.) that was consolidated with Rodriguez. All were eventually designated as class representatives. McGuireWoods LLP was appointed class counsel.

The operative complaint alleges that BAR/BRI has been the major provider of bar preparation courses throughout the United States for decades. In 1995, West started a business called West Bar Review that competed with BAR/BRI in the market for state bar preparation courses. Thomson Company acquired West in 1996 and sought to divest itself of West Bar Review. Kaplan entered into a letter of intent to acquire West Bar Review by early August 1997. BAR/BRI, unaffiliated with West or Kaplan at that time, allegedly sought to thwart the sale of West Bar Review to Kaplan by entering a market division agreement with Kaplan whereby BAR/BRI agreed to pay Kaplan and to withdraw from markets for other test preparation courses, while Kaplan agreed not to enter BAR/BRI's primary market through acquisition of West Bar Review. BAR/BRI then acquired West Bar Review in the fall of 1997. A few years later, in 2001, West bought BAR/BRI.

The pleading also alleges that BAR/BRI erected and maintained various entry barriers to the market for bar preparation courses that included targeting first-year law students with a non-refundable option for BAR/BRI's course when they graduate; offering free access to its Westlaw service to students enrolled in a BAR/BRI course; and advertising constantly on Westlaw, which often has a captive audience of law students required to use the service. It avers that BAR/BRI engaged in numerous other acts of anticompetitive conduct such as entering into an agreement that eliminated a competitor in New York (Marino Bar Review); including non-compete clauses in contracts with law school faculty and other staff to prevent them from working for competitors; destroying competitors' advertising; paying fees to law schools for preferable access; offering a purported scholarship program that actually subsidized students considering a competitor's course; and paying Louisiana State University, a BAR/BRI competitor for preparation courses for the Louisiana bar, to discontinue its course.

Claims are stated for violation of section 7 of the Clayton Act, 15 U.S.C. § 18, by West for the acquisition of West Bar Review by BAR/BRI; violation of section 1 of the Sherman Act, 15 U.S.C. § 1, by West and Kaplan for their market division agreement; and violation of section 2 of the Sherman Act, 15 U.S.C. § 2, by West for BAR/BRI's anticompetitive conduct. The class seeks recovery of actual damages of at least $300 million ($1,000 for each of the estimated 300,000 members) for each claim, treble damages, and injunctive relief.

On May 15, 2006, the district court certified a Fed.R.Civ.P. 23(b)(3) class1 consisting of all persons who purchased a bar review course from BAR/BRI in the United States from August 1997 to the present. West and Kaplan sought interlocutory review of the certification order, which we declined to allow. A notice of class certification was sent to putative class members in the summer of 2006, informing them of their right to opt-out of the class before August 13, 2006.

During discovery, class counsel reviewed more than 400,000 pages of documents, deposed fourteen fact witnesses, and took one deposition pursuant to Fed.R.Civ.P. 30(b)(6). West and Kaplan deposed the seven class representatives and three non-party witnesses. The parties also conducted depositions of five expert witnesses.

Kaplan moved for summary judgment, which the district court denied. West had yet to file a motion for summary judgment when a settlement was reached.

Settlement negotiations began in November 2006. The sessions were mediated by the Honorable Daniel Weinstein (Ret. Superior Court of California, San Francisco County) of JAMS, a provider of ADR services, who had extensive experience in this type of case. An agreement was executed on February 2, 2007, though class representatives Rodriguez, Nesci, and Gintz objected and refused to authorize execution on their behalf. Under the agreement, West and Kaplan agreed to pay $49 million into a settlement fund. The fund is to be allocated pro rata to class members based on the amount each paid for BAR/BRI courses relative to the amounts paid by all other class members who file an allowable claim. Each class member's allocation from the settlement fund is capped at thirty percent of the amount that member paid to BAR/BRI. Any remaining funds are to be distributed by the district court through application of the cy pres doctrine. West and Kaplan also agreed to terminate their marketing agreement; West agreed to include on the form used to enroll law students a statement that an initial payment to BAR/BRI is not a commitment to full payment; and West agreed "that it is committed to accurate advertising as required by the Lanham Act, the Federal Trade Commission Act and similar laws, regulations and rules." In exchange, class members agreed to release all claims against West and Kaplan related to the conduct alleged in the complaint. In Judge Weinstein's opinion, the settlement "was arrived at through arm's length negotiations by counsel who were skilled and knowledgeable about the facts and law of this case," and it was "fair, reasonable and adequate in light of the strengths and weaknesses of the claims and defenses and the risks of establishing liability and damages."

On March 26, 2007, over the objections of Rodriguez, Nesci, and Gintz, the district court granted preliminary approval of the settlement and directed that notice of the settlement be sent to the class, defined as all persons who purchased a bar review course from BAR/BRI anywhere in the United States anytime from August 1, 1997 through July 31, 2006. The Settlement Notice informed class members of the $49 million settlement fund, the methodology for allocating the fund, the non-monetary relief, class counsel's intent to seek twenty-five percent of the settlement fund for attorney's fees and expenses, and counsel's plan to request incentive awards of $25,000 for class representatives Brazeal, Brewer, Frailich, and Rimson, and $75,000 for class representatives Gintz, Nesci, and Rodriguez. It indicated that a final settlement hearing would be held on June 18, 2007, and that the deadline for filing objections was May 21, 2007. The Settlement Notice was mailed to 376,301 people and published in several national periodicals. Fifty-four objections were filed.

Class counsel filed a motion seeking incentive awards for the class representatives after preliminary approval of the settlement and dissemination of the Settlement Notice, but before the final fairness hearing. It turns out that, as part of their retainer agreement, the named plaintiffs in Rodriguez (Rodriguez, Frailich, Nesci, Brazeal, and Gintz) had entered into an incentive arrangement with Van Etten Suzumoto & Becker, which preceded McGuireWoods LLP. The incentive agreements obligated class counsel to seek payment for each of these five in an amount that slid with the end settlement or verdict amount: if the amount were greater than or equal to $500,000, class counsel would seek a $10,000 award for each of them; if it were $1.5 million or more, counsel would seek a $25,000 award; if it were $5 million or more, counsel would seek $50,000; and if it were $10 million or more, counsel would seek $75,000. Neither Brewer nor Rimson, the other two class representatives, was party to an incentive agreement. They were separately represented, by Finkelstein Thompson LLP, and Zwerling, Shachter & Zwerling, LLP. By the time the motion was filed, Brazeal and Frailich had agreed to lower their request to $25,000 from the $75,000 promised in the incentive agreement, but Gintz, Nesci, and Rodriguez, who objected to the settlement, did not. Incentive awards in the amount of $25,000 were also sought for Brewer and Rimson.

A final hearing on the fairness, reasonableness, and adequacy of the settlement was held on June 18 and July 9, 2007. Twelve groups of Objectors questioned the failure to bifurcate the section 7 claim, adequacy of the monetary portion of the settlement, not providing for the break-up of BAR/BRI, the cy pres award, and sealing of documents pursuant to a protective order. On September 10, 2007, the district court gave final approval to the settlement agreement in a thirty-seven page Settlement Order and fifty pages of Findings of Fact and Conclusions of Law. It found that the Settlement Notice and dissemination were adequate, and that the settlement was fair, adequate, and reasonable despite the conflict of interest between class representatives and class members. The court denied the motion for incentive awards to all seven class representatives, finding that the amount was unreasonable in light of the work and risk undertaken, and that the incentive agreements created actual conflicts of interest in violation of public policy. It denied fees to Objectors' counsel because they "did not add anything to the court's order denying" the motion for incentive awards, but awarded class counsel its lodestar, enhanced by a 1.75 multiplier, up to a limit of twenty-five percent of the settlement fund.

Six groups of Objectors have timely appealed.2

II

Much of the appeal turns on the presence — and nondisclosure to the class — of the incentive agreements. In particular, Objectors assert that the Settlement Notice offends due process because it omitted material information about the agreements, and that the settlement itself should have been rejected because the incentive agreements prevented the class representatives from providing adequate representation. Relatedly, Objectors contend that they benefitted the class by successfully opposing the incentive awards and should be allowed attorney's fees for the effort. Conversely, they question the attorney's fee award to class counsel as California law precludes the recovery of fees for conflicting representation.

A

Incentive awards are fairly typical in class action cases. See 4 William B. Rubenstein et al., Newberg on Class Actions § 11:38 (4th ed.2008); Theodore Eisenberg & Geoffrey P. Miller, Incentive Awards to Class Action Plaintiffs: An Empirical Study, 53 U.C.L.A. L.Rev. 1303 (2006) (finding twenty-eight percent of settled class actions between 1993 and 2002 included an incentive award to class representatives). Such awards are discretionary, see In re Mego Fin. Corp. Sec. Litig., 213 F.3d 454, 463 (9th Cir.2000), and are intended to compensate class representatives for work done on behalf of the class, to make up for financial or reputational risk undertaken in bringing the action, and, sometimes, to recognize their willingness to act as a private attorney general. Awards are generally sought after a settlement or verdict has been achieved.

The incentive agreements entered into as part of the initial retention of counsel in this case, however, are quite different. Although they only bound counsel to apply for an award, thus leaving the decision whether actually to make one to the district judge, these agreements tied the promised request to the ultimate recovery and in so doing, put class counsel and the contracting class representatives into a conflict position from day one.

The arrangement was not disclosed when it should have been and where it was plainly relevant, at the class certification stage. Had it been, the district court would certainly have considered its effect in determining whether the conflicted plaintiffs — Rodriguez, Frailich, Nesci, Brazeal, and Gintz — could adequately represent the class. The conflict might have been waived, or otherwise contained, but the point is that uncovering conflicts of interest between the named parties and the class they seek to represent is a critical purpose of the adequacy inquiry. See Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 625, 117 S.Ct. 2231, 138 L.Ed.2d 689 (1997). "[A] class representative must be part of the class and `possess the same interest and suffer the same injury' as the class members." E. Tex. Motor Freight Sys. Inc. v. Rodriguez, 431 U.S. 395, 403, 97 S.Ct. 1891, 52 L.Ed.2d 453 (1977) (quoting Schlesinger v. Reservists Comm. to Stop the War, 418 U.S. 208, 216, 94 S.Ct. 2925, 41 L.Ed.2d 706 (1974)); see also Amchem Prods., 521 U.S. at 625-26, 117 S.Ct. 2231. An absence of material conflicts of interest between the named plaintiffs and their counsel with other class members is central to adequacy and, in turn, to due process for absent members of the class. Hanlon v. Chrysler Corp., 150 F.3d 1011, 1020 (9th Cir.1998).

In fact, the incentive agreements came to the fore when Objectors pounced on them in opposing class counsel's motion for incentive awards to the class representatives. This happened after preliminary approval of the settlement. In that context the district court held that the agreements were inappropriate and contrary to public policy for a number of reasons: they obligate class counsel to request an arbitrary award not reflective of the amount of work done, or the risks undertaken, or the time spent on the litigation; they create at least the appearance of impropriety; they violate the California Rules of Professional Conduct prohibiting fee-sharing with clients and among lawyers; and they encourage figurehead cases and bounty payments by potential class counsel. The court found it particularly problematic that the incentive agreements correlated the incentive request solely to the settlement or litigated recovery, as the effect was to make the contracting class representatives' interests actually different from the class's interests in settling a case instead of trying it to verdict, seeking injunctive relief, and insisting on compensation greater than $10 million. It further observed that the parties' failure to disclose their agreement to the court, and to the class, violated the contracting representatives' fiduciary duties to the class and duty of candor to the court.

We agree. By tying their compensation — in advance — to a sliding scale based on the amount recovered, the incentive agreements disjoined the contingency financial interests of the contracting representatives from the class. As the district court observed, once the threshold cash settlement was met, the agreements created a disincentive to go to trial; going to trial would put their $75,000 at risk in return for only a marginal individual gain even if the verdict were significantly greater than the settlement. The agreements also gave the contracting representatives an interest in a monetary settlement, as distinguished from other remedies, that set them apart from other members of the class. Further, agreements of this sort infect the class action environment with the troubling appearance of shopping plaintiffships. If allowed, ex ante incentive agreements could tempt potential plaintiffs to sell their lawsuits to attorneys who are the highest bidders, and vice-versa. In addition, these agreements implicate California ethics rules that prohibit representation of clients with conflicting interests.3 See Image Tech. Serv., Inc. v. Eastman Kodak Co., 136 F.3d 1354, 1358 (9th Cir.1998) (noting that "[s]imultaneous representation of clients with conflicting interests (and without informed written consent) is an automatic ethics violation in California"); Flatt v. Superior Court, 9 Cal.4th 275, 36 Cal.Rptr.2d 537, 885 P.2d 950, 955 (1994).

Although we have not previously encountered incentive agreements, we expressed concern about similar problems with incentive awards in Staton v. Boeing Co., 327 F.3d 938 (9th Cir.2003). There, we declined to approve a settlement agreement where the awards request indicated that the class representatives were "more concerned with maximizing [their own] incentives than with judging the adequacy of the settlement as it applies to class members at large." See Staton, 327 F.3d at 977-78. We explained that excess incentive awards may put the class representative in a conflict with the class and present a "considerable danger of individuals bringing cases as class actions principally to increase their own leverage to attain a remunerative settlement for themselves and then trading on that leverage in the course of negotiations." Id. at 976-77.4 The danger is exacerbated if the named plaintiffs have an advance guarantee that a request for a relatively large incentive award will be made that is untethered to any service or value they will provide to the class.

In sum, we disapprove of the incentive agreements entered into between the named plaintiffs and class counsel in this case. They created an unacceptable disconnect between the interests of the contracting representatives and class counsel, on the one hand, and members of the class on the other. We expect those interests to be congruent. See Molski v. Gleich, 318 F.3d 937, 955 (9th Cir.2003) (noting that adequate representation consists of an "absence of antagonism" and a "sharing of interests between representatives and absentees") (internal quotation marks and citation omitted). They also gave rise to a disturbing appearance of impropriety. And failing to disclose the incentive arrangements in connection with class certification compounded these problems by depriving the court, and the class, of the safeguard of informed judicial consideration of the adequacy of class representation.

This said, we do not believe the district court was required to reject the settlement for inadequate representation. Only five of the seven class representatives had an incentive agreement. Brewer and Rimson did not. "[T]he adequacy-of-representation requirement is satisfied as long as one of the class representatives is an adequate class representative." Local Joint Executive Bd. of Culinary/Bartender Trust Fund v. Las Vegas Sands, Inc., 244 F.3d 1152, 1162 n. 2 (9th Cir.2001); 7A Charles A. Wright, Arthur R. Miller & Mary Kay Kane, Federal Practice and Procedure § 1765, at 326 (2005) ("[I]f there is more than one named representative, it is not necessary that all the representatives meet the Rule 23(a)(4) standard; as long as one of the representatives is adequate, the requirement will be met."). Brewer and Rimson were also separately represented. There is no evidence or contention that these two had any other conflict with the class.

Objectors submit only that Brewer and Rimson should have told the district court about the incentive agreements. Even assuming these two knew about the incentive agreements and understood the implications, this did not create a conflict of interest or otherwise interfere with their ability or motivation to represent the class. Other factors indicate that the class was adequately represented: Judge Weinstein, who mediated the settlement, attested that the negotiations were conducted at arm's length; there was no evidence of collusion;5 and the settlement fund far exceeded the ten million dollar trigger for the contracting class representatives' incentive agreements.

Accordingly, we conclude that the presence of conflicted representatives was harmless. Similarly, the adequacy requirement for class counsel is satisfied. Fed.R.Civ.P. 23(a)(4), (g)(4). Class counsel vigorously prosecuted the case through to a fair settlement with the participation of two nonconflicted law firms that represented class representatives Brewer and Rimson. See Hanlon, 150 F.3d at 1020. The court found their representation was adequate; and Judge Weinstein, who oversaw the settlement negotiations, believed that "[e]ach side aggressively advocated their positions," class counsel "ha[d] as their primary goal achieving the maximum substantive relief that they could," and agreement "was arrived at through arm's length negotiations by counsel who were skilled and knowledgeable about the facts and law of this case." Moreover, the participation of two firms that did not enter incentive agreements, Finkelstein Thompson and Zwerling, Shachter & Zwerling, assuages any additional concerns that a conflict created by the incentive agreements may have adversely affected the adequacy of representation. See Linney v. Cellular Alaska P'ship, 151 F.3d 1234, 1239 (9th Cir.1998) ("[T]he addition of new and impartial counsel can cure a conflict of interest even where previous counsel continues to be involved in the case.").6

B

It follows that the Settlement Notice was not fatally defective for failing to disclose the actual or potential conflict arising out of the existence of incentive agreements.

"The court must direct notice in a reasonable manner to all class members who would be bound by the proposal." Fed. R.Civ.P. 23(e)(1). "Notice is satisfactory if it `generally describes the terms of the settlement in sufficient detail to alert those with adverse viewpoints to investigate and to come forward and be heard.'" Churchill Vill., LLC v. Gen. Elec., 361 F.3d 566, 575 (9th Cir.2004) (quoting Mendoza v. Tucson Sch. Dist. No. 1, 623 F.2d 1338, 1352 (9th Cir.1980)).

The Settlement Notice advised absent class members that an application would be made to the court for an incentive award of $25,000 for Frailich, Brazeal, Brewer, and Rimson, and $75,000 for Rodriguez, Nesci, and Gintz to compensate them for their participation in, and prosecution of, this case on behalf of the class, and that the petition for incentive awards, which would be filed before May 7, 2007, would be available for inspection at the clerk's office. Rodriguez, Nesci, and Gintz disclosed the existence of the incentive agreements in a document filed with the court on March 6, 2007. The Notice also indicated that the settlement agreement and related documents were posted at www.barbri-classaction.com, and provided class counsel's phone number and an email address to which inquires could be sent. The settlement agreement itself states that whether incentive awards should be awarded to class representatives, as well as attorney's fees, will be determined at the final settlement hearing. It also provides that class counsel will submit an application for an incentive award to each class representative to be paid from the gross settlement fund, and that West and Kaplan agree not to oppose any application for an incentive award seeking no more than $25,000.

While neither the Settlement Notice nor the settlement agreement discloses the incentive agreements, both show that incentive awards will be sought. In the circumstances this was sufficient to alert class members to follow-up if they had concerns. See id.

Objectors contend that the Settlement Notice also failed to provide a meaningful description of the terms of the settlement, including the content of objections and the expected value of fully litigating the case. In our view, the Notice contains adequate information, presented in a neutral manner, to apprise class members of the essential terms and conditions of the settlement. The Notice advises class members that a majority (hence, not all) of the class representatives approve the settlement. It describes the aggregate amount of the settlement fund and the plan for allocation, thereby complying with what we require. See Torrisi v. Tucson Elec. Power Co., 8 F.3d 1370, 1373-74 (9th Cir.1993); Marshall v. Holiday Magic, Inc., 550 F.2d 1173, 1177-78 (9th Cir.1977). While the Notice does not detail the content of objections, or analyze the expected value, we do not see why it should. Settlement notices are supposed to present information about a proposed settlement neutrally, simply, and understandably7 — objectives not likely served by including the adversarial positions of objectors. We therefore conclude that the Notice communicated the essentials of the proposed settlement in a sufficiently balanced, accurate, and informative way to satisfy due process concerns.

C

Objectors who challenged the incentive awards argue that the district court improperly denied fees attributable to that work. The court rejected their request for fees on the footing that Objectors' counsel "did not add anything" to its decision to deny incentive awards. This seems clearly erroneous to us. The court was not focused on the incentive agreements before Objectors took exception to them after the motion to award payments to the class representatives was filed. In the wake of that objection, the court denied the motion for incentive awards in its entirety because the amounts requested were unreasonable and the incentive agreements were inappropriate and contrary to public policy. The net effect was to leave $325,000 in the settlement fund — for distribution to the class as a whole — that otherwise would have gone to the class representatives. Given this, we cannot let stand a ruling that Objectors did nothing that increased the fund or substantially benefitted the class members. See Vizcaino v. Microsoft Corp., 290 F.3d 1043, 1051 (9th Cir.2002). Therefore, we remand for the district court to reconsider the extent to which Objectors added value that increased the fund or substantially benefitted the class members, and to award attorney's fees accordingly.

III

Objectors press a number of issues apart from the effect of the incentive agreements that bear on whether the settlement was fair, adequate, and reasonable. The most serious have to do with the district court's evaluation of the amount offered in settlement, in particular, its failure to estimate the range of possible outcomes and ascribe a probability to those outcomes, and to consider treble damages.

Fed.R.Civ.P. 23(e) requires judicial approval of any settlement by a certified class. The settlement must be "fair, reasonable, and adequate." Fed.R.Civ.P. 23(e)(2). A district court "may consider some or all of the following factors" when assessing whether a class action settlement agreement meets this standard:

Molski, 318 F.3d at 953; accord Staton, 327 F.3d at 959.

We review approval of a class action settlement for a "clear abuse of discretion." Molski, 318 F.3d at 953. This court "`will affirm if the district judge applies the proper legal standard and his or her findings of fact are not clearly erroneous.'" Id. (quoting In re Mego Fin. Corp. Sec. Litig., 213 F.3d at 458). "Our review of the district court's decision to approve a class action settlement is extremely limited. It is the settlement taken as a whole, rather than the individual component parts, that must be examined for overall fairness." Hanlon, 150 F.3d at 1026 (internal citation omitted). "To survive appellate review, the district court must show it has explored comprehensively all factors," id., but a court is not required to "reach any ultimate conclusions on the contested issues of fact and law which underlie the merits of the dispute, for it is the very uncertainty of outcome in litigation and avoidance of wasteful and expensive litigation that induce consensual settlements," Officers for Justice v. Civil Serv. Comm'n of San Francisco, 688 F.2d 615, 625 (9th Cir.1982).

Here, the court balanced each of the relevant factors in approving the settlement.

Strength of the plaintiffs' case. The court noted that successfully opposing Kaplan's motion for summary judgment did not mean that the class had established liability or would obtain a favorable, unanimous jury verdict. This is, of course, correct. It noted the difficulty of proving an antitrust case, and that Kaplan and West had substantive and procedural defenses to all three of the class's claims (including a potential statute of limitations defense that could decrease the size of the class). Also, there were no government coattails for the class to ride. Counting this factor in favor of settlement was not a clear abuse of discretion.

Amount offered in settlement. Objectors claim it was legal error for the court to consider only estimates of single damages without considering the treble damages that are an automatic component of the recovery of antitrust damages. 15 U.S.C. § 15(a). The district court found that the $49 million settlement represented thirty percent of the damages estimated by the class expert — $158 million to $168 million. This analysis compared the settlement amount to the best possible outcome for the class, without taking into account the significant difference between the class's estimate and the defense's. The defense expert had opined that there likely would be no damages but if there were any, they would not exceed $7 million. In any event, the court declined to accept Objectors' argument that the monetary portion of the settlement was inadequate because the section 7 claim was worth treble the class expert's single damages estimate — or $360 million. It reasoned that doing so would presuppose that plaintiffs prevail at the end of trial (thus undercutting the point of a negotiated resolution where defendants do not admit liability); it would be speculative; and, as the Second Circuit indicated in City of Detroit v. Grinnell Corp., 495 F.2d 448, 458 (2d Cir.1974), overruled on other grounds as recognized by U.S. Football League v. Nat'l Football League, 887 F.2d 408, 415-16 (2d Cir.1989), courts do not traditionally factor treble damages into the calculus for determining a reasonable settlement value.

It is our impression that courts generally determine fairness of an antitrust class action settlement based on how it compensates the class for past injuries, without giving much, if any, consideration to treble damages.8 At the same time, treble damages are a fact of life in antitrust litigation. In some cases a court, asked to approve a settlement, may believe the class's claim is so strong that the merits of the amount negotiated cannot reasonably be evaluated without measuring it against the likelihood of a treble as well as a single damages recovery. We have never precluded courts from comparing the settlement amount to both single and treble damages. By the same token, we do not require them to do so in all cases.

This circuit has long deferred to the private consensual decision of the parties. See Hanlon, 150 F.3d at 1027. Experienced counsel such as those representing all the parties in this case will certainly be aware of exposure to treble damages in an antitrust action. Likewise, the mediator in this case. As we have emphasized,

Id. (quoting Officers for Justice, 688 F.2d at 625).

In this case, the negotiated amount is fair and reasonable no matter how you slice it. There is no evidence of fraud, overreaching, or collusion. Even considering the trebling effect, the settlement amount represents approximately ten percent of the class's estimate of its own trebled damages and more than twice that estimated by West and Kaplan. The $49 million is in cash, not in kind, which is a good indicator of a beneficial settlement. All things considered, the district court neither committed legal error, nor aside from that, clearly abused its discretion in weighing the amount offered in settlement in favor of approving the settlement.

We are not persuaded otherwise by Objectors' further submission that the court should have specifically weighed the merits of the class's case against the settlement amount and quantified the expected value of fully litigating the matter. For this they rely on the Seventh Circuit's opinion in Synfuel Tech., Inc. v. DHL Express (USA), Inc., 463 F.3d 646 (7th Cir. 2006), which follows that circuit's precedent requiring district courts to determine the strength of the plaintiff's case on the merits balanced against the amount offered in settlement by "`quantifying the net expected value of continued litigation to the class.'" Id. at 653 (quoting Reynolds v. Beneficial Nat'l Bank, 288 F.3d 277, 284-85 (7th Cir.2002)). To do this, the Seventh Circuit directs courts to "`estimate the range of possible outcomes and ascrib[e] a probability to each point on the range.'" Id. However, our approach, and the factors we identify, are somewhat different. We put a good deal of stock in the product of an arms-length, non-collusive, negotiated resolution, Hanlon, 150 F.3d at 1027; Officers for Justice, 688 F.2d at 625, and have never prescribed a particular formula by which that outcome must be tested. As we explained in Officers for Justice, "[u]ltimately, the district court's determination is nothing more than an amalgam of delicate balancing, gross approximations and rough justice." 688 F.2d at 625 (internal quotation marks and citation omitted). The Seventh Circuit also recognizes that precision is impossible, and that even its more structured approach is apt to produce only a "ballpark valuation." Synfuel, 463 F.3d at 653.

In reality, parties, counsel, mediators, and district judges naturally arrive at a reasonable range for settlement by considering the likelihood of a plaintiffs' or defense verdict, the potential recovery, and the chances of obtaining it, discounted to present value. See Federal Judicial Center, Manual for Complex Litigation § 21.62, at 316 (4th ed.2004) (one factor "that may bear on review of a settlement" is "the advantages of the proposed settlement versus the probable outcome of a trial on the merits of liability and damages as to the claims, issues, or defenses of the class and individual class members"); In re Gen. Motors Corp. Pick-Up Truck Fuel Tank Prods. Liab. Litig., 55 F.3d 768, 806 (3d Cir.1995).

Although the district court did not put it this way, the amount of the alleged overcharge, the estimated recovery ranges by both parties and their experts, and the results of a mediated resolution, were before it. Objectors do not explain how reversing the math on the record would have yielded a meaningfully different result. Accordingly, the court did not clearly abuse its discretion in concluding that this factor weighs in favor of approving the settlement.

In an argument related to their position on treble damages, Objectors also challenge the cy pres provision, which they point out is a disfavored substitute for distribution of benefits directly to class members. See Molski, 318 F.3d at 954-55. Here, their argument goes, the settlement agreement distributes funds to class members up to thirty percent of the amount each paid to BAR/BRI; this is based on an estimate of single damages; thus, the cy pres provision effectively substitutes for treble damages that should also be distributed to class members. However, this issue becomes ripe only if the entire settlement fund is not distributed to class members. See Six (6) Mexican Workers v. Ariz. Citrus Growers, 904 F.2d 1301, 1313 (9th Cir.1990) (Fernandez, J., concurring). That trigger point has not been reached; no cy pres disbursement is imminent; and the fund in this case may well be depleted before cy pres kicks in. We therefore decline to consider the propriety of cy pres at this time.

Risk, expense, complexity, and likely duration of further litigation. The court found, with substantial support in the record, that the case is complex and likely to be expensive and lengthy to try. The class in this case does not have the benefit, like some other antitrust classes, of previous litigation between the defendants and the government. While Objectors point out that much heavy-lifting had already been done, a number of serious hurdles remained — Daubert motions,9 West's anticipated motion for summary judgment, and a motion to bifurcate. Inevitable appeals would likely prolong the litigation, and any recovery by class members, for years. This factor, too, favors the settlement.

Risk of maintaining class action status. The court did not have to analyze the probabilities that West and Kaplan would seek decertification of the nationwide class and succeed in the endeavor, as Objectors suggest, to weigh this factor in favor of the settlement. A district court may decertify a class at any time. See Gen. Tel. Co. of Sw. v. Falcon, 457 U.S. 147, 160, 102 S.Ct. 2364, 72 L.Ed.2d 740 (1982). West and Kaplan vigorously opposed certification of a nationwide class, sought (albeit unsuccessfully) to take an interlocutory appeal from that order, and would undoubtedly have appealed certification if there were a final, adverse judgment. At the time of settlement, the risk remained that the nationwide class might be decertified; it was not so minimal that this factor could not weigh in favor of the settlement. Bar review courses are given on a state-by-state basis; states are distinct markets geographically, and possibly in other respects. This could call the nationwide class into question.

Discovery completed; state of proceedings. Extensive discovery had been conducted, and the parties had gone through one round of summary judgment proceedings. From this the district court could find that counsel had a good grasp on the merits of their case before settlement talks began. Nor is there any dispute that counsel had considerable experience in litigating antitrust matters, class actions, and other complex litigation. As we have held that "[p]arties represented by competent counsel are better positioned than courts to produce a settlement that fairly reflects each party's expected outcome in litigation," In re Pac. Enters. Sec. Litig., 47 F.3d 373, 378 (9th Cir.1995), the court could weigh this factor in favor of approval.

Reaction to proposed settlement. The court had discretion to find a favorable reaction to the settlement among class members given that, of 376,301 putative class members to whom notice of the settlement had been sent, 52,000 submitted claims forms and only fifty-four submitted objections. See, e.g., Churchill Village, 361 F.3d at 577 (affirming approval of a class action settlement where forty-five objections were received out of 90,000 notices).10

For these reasons, we cannot say that the district court clearly abused its discretion in approving the settlement.

IV

Having upheld approval of the settlement agreement, we must consider the award of attorney's fees to class counsel. This brings us back to the incentive agreements.

We require only that fee awards be reasonable in the circumstances, In re Wash. Pub. Power Supply Sys. Sec. Litig., 19 F.3d 1291, 1294 n. 2 (9th Cir.1994), and our review is for abuse of discretion, Powers v. Eichen, 229 F.3d 1249, 1256 (9th Cir.2000). The district court may award fees pursuant to either a lodestar or a straight percentage of the settlement fund. Id. (quoting In re Coordinated Pretrial Proceedings in Petroleum Prods. Antitrust Litig., 109 F.3d 602, 607 (9th Cir.1997)). Here it adopted the lodestar. A court may also apply a multiplier to the lodestar calculation, which the district court did. Wash. Pub. Power Supply Sys. Sec. Litig., 19 F.3d at 1299-1301. In doing so it found that counsel faced substantial risk in prosecuting this action; did not have the benefit of fruits from underlying government actions; there were no controlling precedents, especially with regard to the section 7 claim; defense counsel were skilled and formidable; and there were a number of hurdles in proving both damages and liability. The problem is that the district court nowhere appears to have considered the effect on the award of attorney's fees of the conflict of interest that resulted from the incentive agreements.

By virtue of the district court's local rules, California law controls whether an ethical violation occurred. C.D. Cal. L.R. 83-3.1.2. "Simultaneous representation of clients with conflicting interests (and without written informed consent) is an automatic ethics violation in California and grounds for disqualification." Image Tech. Serv., 136 F.3d at 1358; Flatt, 36 Cal.Rptr.2d 537, 885 P.2d at 955. Under California law, "[a]n attorney cannot recover fees for such conflicting representation." Image Tech. Serv., 136 F.3d at 1358. "An attorney may claim fees only for services provided before the conflict arose and the ethical breach occurred." Id.

We express no opinion on the impact of these principles on the fees request in this case, but it is apparent that they are, at least, implicated. We realize that conflicts of interest among class members are not uncommon and arise for many different reasons. However, the conflict of interest inhering in the incentive agreements did not just happen, nor was it a conflict that developed beyond the control or perception of class counsel. It was inserted into the retainer agreement. "`The responsibility of class counsel to absent class members whose control over their attorneys is limited does not permit even the appearance of divided loyalties of counsel.'" Kayes v. Pac. Lumber Co., 51 F.3d 1449, 1465 (9th Cir.1995) (quoting Sullivan v. Chase Inv. Servs. of Boston, Inc., 79 F.R.D. 246, 258 (N.D.Cal.1978)). In addition, class counsel's fiduciary duty is to the class as a whole and it includes reporting potential conflict issues. Neither the incentive agreements nor the possibility of conflict was disclosed to the court so that it could take steps to protect the interests of absentee class members. We think it appropriate for the district court to consider whether counsel could represent both the class representatives with whom there was an incentive agreement, and absentee class members, without affecting the entitlement to fees.

At the fee-setting stage when fees are to come out of the settlement fund, the district court has a fiduciary role for the class. See Wash. Pub. Power Supply Sys. Sec. Litig., 19 F.3d at 1302. It may be that the record is insufficient for the court to make a reasoned judgment; if so, an opportunity should be afforded for the parties to develop the record. Accordingly, we remand for the district court to consider in the first instance the effect, if any, of the conflict arising out of the incentive agreements on the request by class counsel for an attorney's fee award.

Objectors argue the amount of the fee award to class counsel, including the 1.75 multiplier and the cap at twenty-five percent of the settlement fund, is grossly excessive. We decline to address this argument at this time. On remand, we expect the district court to revisit all aspects of the award to class counsel.

V

We conclude that incentive agreements, entered into as part of five named plaintiffs' retainer agreement with counsel, created conflicts among them (later certified as class representatives), their counsel (later certified as class counsel), and the rest of the class. It was inappropriate not to disclose these agreements at the class certification stage, because an ex ante incentive agreement is relevant to whether a named plaintiff who is party to one can adequately represent the class. However, this impropriety did not require the district court to reject the settlement negotiated in this case because two nonconflicted class representatives with non-conflicted counsel participated. Nor did the district court clearly abuse its discretion in determining that the amount of the settlement favored approval, whether compared to the likely recovery of single, or treble, damages. By any measure, this settlement is fair, adequate, and reasonable.

The district court should have recognized that Objectors' position on the impropriety of incentive agreements had some effect on its decision to deny the request for incentive awards; and it should have considered what effect, if any, the ethics implications of a conflict of interest created by the incentive agreements had on class counsel's request for an award of attorney's fees.

Therefore, we affirm approval of the settlement. We reverse and remand the award of attorney's fees to class counsel for consideration of the effect, if any, of the incentive agreements on entitlement to fees. We also reverse and remand the denial of fees to Objectors' counsel for a determination of a reasonable amount given their contribution to the denial of the requests for incentive awards.

AFFIRMED IN PART; REVERSED AND REMANDED IN PART.

Each party shall bear its own costs on appeal.